Back to EveryPatent.com
United States Patent | 5,742,740 |
McCormack ,   et al. | April 21, 1998 |
A method of operating an adaptive, or neural, network is disclosed for performing first break analysis for seismic shot records. The adaptive network is first trained according to the generalized delta rule. The disclosed training method includes selection of the seismic trace with the highest error, where the backpropagation is performed according to the error of this worst trace. The learning and momentum factors in the generalized delta rule are adjusted according to the value of the worst error, so that the learning and momentum factors increase as the error decreases. The training method further includes detection of slow convergence regions, and methods for escaping such regions including restoration of previously trimmed dormant links, renormalization of the weighting factor values, and the addition of new layers to the network. The network, after the addition of a new layer, includes links between nodes which skip the hidden layer. The error value used in the backpropagation is reduced from that actually calculated, by adjusting the desired output value, in order to reduce the growth of the weighting factors. After the training of the network, data corresponding to an average of the graphical display of a portion of the shot record, including multiple traces over a period of time, is provided to the network. The time of interest of the data is incremented until such time as the network indicates that the time of interest equals the first break time. The analysis may be repeated for all of the traces in the shot record.
Inventors: | McCormack; Michael D. (Plano, TX); Rock; Alan D. (Knoxville, TN) |
Assignee: | Atlantic Richfield Company (Los Angeles, CA) |
Appl. No.: | 007902 |
Filed: | January 15, 1993 |
Current U.S. Class: | 706/14; 706/16; 706/25 |
Intern'l Class: | G06F 015/18 |
Field of Search: | 395/22,23,24 |
3602888 | Aug., 1971 | Nishiyama | 395/24. |
3638196 | Jan., 1972 | Nishiyama | 395/24. |
3950733 | Apr., 1976 | Cooper et al. | 395/24. |
4101862 | Jul., 1978 | Martin | 367/54. |
4236233 | Nov., 1980 | Davis, Jr. et al. | 367/38. |
4609471 | Sep., 1986 | Silverman | 367/54. |
4663743 | May., 1987 | Rampuria et al. | 367/68. |
4692909 | Sep., 1987 | Gard et al. | 364/421. |
4695984 | Sep., 1987 | Paal | 367/36. |
4725992 | Feb., 1988 | McNatt et al. | 364/421. |
4876673 | Oct., 1989 | McCowan | 364/421. |
4876731 | Oct., 1989 | Loris et al. | 395/21. |
4912647 | Mar., 1990 | Wood | 395/23. |
4912649 | Mar., 1990 | Wood | 395/23. |
4912651 | Mar., 1990 | Wood et al. | 395/23. |
4912652 | Mar., 1990 | Wood | 395/23. |
4912653 | Mar., 1990 | Wood | 395/23. |
4912654 | Mar., 1990 | Wood | 395/23. |
4912655 | Mar., 1990 | Wood | 395/23. |
4914603 | Apr., 1990 | Wood | 395/23. |
4933872 | Jun., 1990 | Vandenberg et al. | 395/22. |
4964087 | Oct., 1990 | Widrow | 364/421. |
4984220 | Jan., 1991 | Bodine et al. | 364/421. |
4985873 | Jan., 1991 | Eyl et al. | 367/34. |
5001677 | Mar., 1991 | Masters | 367/68. |
5033006 | Jul., 1991 | Ishizuka et al. | 395/24. |
5050095 | Sep., 1991 | Samad | 395/23. |
5056037 | Oct., 1991 | Eberhardt | 395/23. |
5093899 | Mar., 1992 | Hiraiwa | 395/23. |
5107454 | Apr., 1992 | Niki | 395/24. |
5113483 | May., 1992 | Keeler et al. | 395/24. |
5129039 | Jul., 1992 | Hiraiwa | 395/24. |
5150323 | Sep., 1992 | Castelaz | 395/24. |
5276769 | Jan., 1994 | Ueda et al. | 395/24. |
Coppens, "First Arrival Picking on Common-Offset Trace Collections for Automatic Estimation of Static Corrections", Geophysical Propsecting 33, 1212-31, 1985. Olsen, "A Stable and Flexible Procedure for the Inverse Modelling of Seismic First Arrivals," Geophysical Prospecting 37, 455-65, 1989. Spagnolini, "Adaptive Picking of Refracted First Arrivals," Geophysical Propsecting 39, 293-312, 1991. Veezhinathan, et al., "A Neural Network Approach to First Break Picking", International Joint Conference of Neural Networks, vol. I (San Diego, CA. Jun. 17-21, 1990) pp. 235-240. Gelchinsky, et al., "Automated Picking of First Arrivals and Parameterization of Travel Time Curves", Geophysical Prospecting 31 (1983), pp. 915-928. Ervin, et al., "Automated Analysis of Marine Refraction Data: A Computer Algorithm", Geophysics, vol. 48, No. 5 (May 1983), pp. 582-589. Malferrari, et al., An Application of Neural Networks to Oil Well Drilling, INNC. 90 Paris Internat. Neural Network Conf. v. 1, pp. 127-130. Bitto, et al., "Seismic Event Discrimination Using Neural Networks," 23rd Asilomar Conference on Sumals, Systems and Computers, v. 1, 1989, pp. 326-330. Liv. "A Rule-Based System for Automatic Seismic Discrimination," Pattern Recognition, v. 19, N.6, 1985 pp. 459-463. Caudill, "Neural Networks Primer, Part I", AI Expert (Dec. 1987), pp.46-55. Caudill, "Neural Networks Primer, Part II", AI Expert (Feb. 1988), pp. 55-67. Caudill, "Neural Networks Primer, Part III", AI Expert (Jun. 1988), pp. 53-65. Wasserman, et al., "Neural Networks, Part 1", IEEE Expert, (1987), pp. 10-14. Wasserman, et al., "Neural Networks, Part 2"; IEEE Expert (1988), pp. 10-15. Lippmann, "An Introduction to Computing with Neural Nets", IEEE ASSP Magazine (Apr. 1987), pp. 4-24. Zhu, et al., "Training Neural Networks for FLg Feature Recognition," INNC 90 Internat. Neural Network Conf. v.1, pp. 137-140. Epping, et al., "A Neural Network for Analysis and Improvement of Gas Well Production," Proc. 1990 Summer Computer Simulation Conf. pp. 791-796. |